An international team of astronomers has conducted what is believed to be the largest cosmological computer simulation ever. The simulation, known as FLAMINGO, tracks both dark and ordinary matter in order to gain insight into the evolution of the universe. The simulation calculates the evolution of all the components of the universe, including dark matter, dark energy, and ordinary matter, according to the laws of physics. The results of the simulation have been published in three papers in the Monthly Notices of the Royal Astronomical Society.
Facilities like the Euclid Space Telescope and NASA’s JWST collect vast amounts of data on galaxies, quasars, and stars. Simulations like FLAMINGO play a crucial role in interpreting this data by connecting predictions from theories of the universe to the observed data. The simulation allows scientists to investigate possible biases in measurements, which could help explain discrepancies between different cosmological observations.
One of the most significant findings from the FLAMINGO simulations is that both neutrinos and ordinary matter are crucial for making accurate predictions. Simulations that include ordinary matter are more challenging and require more computing power due to the complex nature of ordinary matter, which feels not only gravity but also gas pressure. Additionally, the contribution of neutrinos, subatomic particles with small but uncertain mass, is also important but has not been simulated until now.
The FLAMINGO project used machine learning to calibrate the effect of galactic winds by comparing predictions from various simulations with observed masses of galaxies and the distribution of gas in galaxy clusters. The researchers conducted simulations using different cosmic volumes, resolutions, and parameters, including the strength of galactic winds and the mass of neutrinos. The largest simulation utilized 300 billion resolution elements and is considered the largest cosmological computer simulation with ordinary matter completed to date.
The FLAMINGO simulations offer a new virtual window into the universe, allowing scientists to optimize cosmological observations. The simulations also enable the exploration of new theoretical discoveries and the testing of data analysis techniques, including machine learning. By comparing virtual universes with large-scale structure observations, astronomers can measure the values of cosmological parameters and uncertainties.
Overall, the FLAMINGO simulations provide valuable insights into the evolution of the universe and offer a platform for further advancements in cosmological research. The simulations bridge the gap between theories of the universe and observed data, allowing scientists to refine their understanding of the cosmos. The published papers detailing the FLAMINGO project provide further information and analysis of the simulation’s findings.